Integrating AI, VR, CAD, and Rapid Prototyping for Multi-Utility Electric Vehicle Cabin Design

The accelerating global transition toward sustainable transportation has intensified the need for advanced electric vehicle (EV) technologies capable of supporting diverse mobility functions across urban and peri-urban regions [1]. For emerging economies such as Indonesia, EV development is not only a strategic component of national electrification policies but also an essential pathway for strengthening technological sovereignty and reducing dependence on imported components [2]. Within this broader agenda, multi-utility EVs differ from conventional passenger EVs because they must support variable passenger-cargo configurations, frequent stop-and-go logistics, public-service deployment, operator comfort during extended work cycles, simplified maintenance, and affordable production for small and medium manufacturers. These requirements make the cabin more than a comfort-oriented passenger compartment; it becomes a work interface that integrates visibility, reachability, cargo access, dashboard robustness, ingress-egress efficiency, and manufacturable modularity [3-6].

Recent literature indicates that the enabling technologies for this task have advanced rapidly but remain unevenly integrated. Virtual simulation tools have improved automotive human-machine interaction and ergonomic assessment by allowing users to evaluate cockpit layout, sightlines, posture, and interface accessibility before physical fabrication [7-11]. Generative AI and AI-assisted (Vizcom AI for Education) design have expanded the ideation space for automotive interiors and intelligent cabins, but most applications remain focused on styling images, seat concepts, or component-level optimization rather than Computer-Aided Design (CAD)-ready full-cabin engineering models [12-17]. Similarly, CAD-additive manufacturing workflows and rapid prototyping have become important for shortening prototype cycles, but they are often positioned as downstream validation tools rather than as active feedback nodes linked with ergonomic and aerodynamic simulation [18-24].

However, current EV design workflows remain constrained by sequential development processes [25], lengthy iteration cycles [26], and high prototyping costs [27]. These challenges are particularly acute for small and medium-scale manufacturers in developing countries, who lack the extensive resources and infrastructure typically available to global automotive firms. Existing attempts to combine digital tools have generally emphasized either model-based EV engineering, CAD-cloud transformation, virtual prototyping, or AI-CAD integration [19, 23, 25, 28-30]. They provide valuable foundations, but few studies explicitly connect image-based generative concept exploration, parametric CAD reconstruction, Virtual Reality (VR) ergonomic testing, Computational Fluid Dynamics (CFD) aerodynamic evaluation, and physical rapid prototyping within one traceable cabin-design workflow for a multi-utility EV. Therefore, the novelty of this study is not the isolated use of AI, VR, CAD, CFD, or prototyping, but the operational integration of these tools into a repeatable feedback ecosystem that supports design decisions under emerging-market manufacturing constraints.

A systematic review of existing literature reveals four fundamental gaps. First, CAD studies overwhelmingly prioritize structural and aerodynamic performance but seldom incorporate early-stage ergonomic or human-centered parameters, causing misalignment between engineering feasibility and user requirements [8, 9]. Second, VR-based ergonomic evaluations, despite their growing relevance, are typically implemented after major design decisions are finalized, limiting their influence on upstream conceptual development [10, 11]. Third, AI-driven generative design remains underutilized in automotive cabin research and is often limited to generative styling or image-based concept exploration rather than direct geometry generation for CAD-ready production [12, 13, 16, 17]. In this study, AI is positioned specifically as image-based generative styling that supports early concept exploration; it does not automatically generate manufacturable CAD geometry. Fourth, rapid prototyping processes commonly function as isolated validation steps rather than integrated nodes within an iterative digital feedback loop [18, 19, 24]. Collectively, these shortcomings highlight the absence of a cohesive digital methodology capable of unifying ergonomic, aesthetic, structural, aerodynamic, and manufacturability considerations in a single continuous design cycle.

Addressing these gaps, this study introduces a unified digital design workflow that integrates AI-assisted generative concept exploration, parametric CAD modelling, VR-based ergonomic evaluation, CFD-based aerodynamic simulation, and rapid prototyping into an iterative, fully interconnected development framework. This workflow is implemented through a multi-year collaboration between Universitas Pembangunan Jaya and PT Spora Teknika Indonesia, supported by a research ecosystem that includes prior micro-EV platforms, hybrid solar-electric carts, and cargobike prototypes [31, 32]. The integrated framework enables real-time feedback between conceptual, ergonomic, structural, aerodynamic, and manufacturability assessments, significantly reducing design errors and compressing development timelines, outcomes that are crucial to enhancing EV innovation capacity in emerging economies.

This study contributes to EV design and engineering literature in four major ways: (1) Methodological contribution: It proposes a novel end-to-end digital workflow that merges AI, VR, CAD, CFD, and rapid prototyping into a continuous co-design cycle, closing the methodological fragmentation prevalent in existing research. (2) Empirical contribution: The study demonstrates the real-world implementation of this workflow through the complete development of a multi-utility EV cabin, involving validated outputs such as ergonomic VR insights, CFD-based aerodynamic performance data, AI-generated design variants, 3D-printed full-scale mockups, and CAD-based structural models. (3) Theoretical contribution: It advances the underexplored domain of ergonomic-aerodynamic co-optimization by embedding VR and CFD analyses into the early design stages, enabling simultaneous improvement of human-machine interaction and energy efficiency. (4) Practical contribution: It offers a scalable and cost-efficient development model tailored to the constraints of emerging-market EV manufacturing, providing actionable guidance for domestic industries seeking to enhance competitiveness and accelerate innovation.

By bridging critical methodological, empirical, and industrial gaps, this research positions integrated digital workflows as a strategic enabler for next-generation EV development. The proposed approach supports not only higher Technology Readiness Level (TRL) progression and manufacturability but also contributes to national initiatives aimed at establishing a robust, competitive, and sustainable electric mobility ecosystem.

The findings demonstrate that the integrated workflow accelerated and strengthened the development of a multi-utility EV cabin by connecting concept generation, CAD reconstruction, VR ergonomic feedback, CFD analysis, and physical validation. The AI-assisted stage generated 12 exterior and interior design alternatives, from which 3 concepts were shortlisted using the weighted criteria of ergonomic feasibility, aerodynamic plausibility, manufacturability, modularity, and visual identity. These automatically generated variations (Figure 1) expanded the creative design space and reduced dependence on sequential manual sketching. The selected concept was not transferred directly into CAD as an AI mesh; rather, it served as a styling and proportion reference for parametric reconstruction.

Figure 1. AI-generated (Vizcom AI for Education) vehicle exterior design iterations

Based on the selected AI-generated concepts, detailed three-dimensional CAD models were developed to represent both the exterior and interior structures of the vehicle. The CAD models captured key design elements such as body contours, dashboard geometry, seating layout, and component integration points (Figure 2 and Figure 3). The parametric modelling approach allowed rapid iteration and continuous refinement, particularly when ergonomic or aerodynamic adjustments were required. This ensured that each modification could be efficiently propagated throughout the model, maintaining dimensional accuracy and design consistency across all components. CAD models also served as the central digital asset for downstream simulations, VR evaluations, and prototyping activities.

Once the digital geometry reached an initial level of maturity, VR-based ergonomic assessment (Figure 4) was conducted by importing the CAD model into an immersive environment. Participant feedback and observation identified three principal ergonomic issues in the baseline version: steering reach, dashboard depth, and forward visibility. Specifically, 2 of 5 participants reported that the steering position required excessive reach, 2 of 5 indicated that the dashboard depth reduced perceived openness, and 2 of 5 noted limited lower-front visibility. The design team responded by adjusting the steering column inclination by 20 degrees, reducing dashboard depth by 10 mm, and modifying the windshield angle by 50 degrees. These before-and-after refinements show how VR feedback translated into measurable CAD changes rather than remaining qualitative comments.

Figure 2. Vehicle Computer-Aided Design (CAD) modeling

Figure 3. Computer-Aided Design (CAD) model of interior (dashboard and seating layout)

Figure 4. VR-based ergonomic evaluation of driver position

Figure 5. Computational Fluid Dynamics (CFD) simulation showing pressure distribution across the body surface

To complement the ergonomic evaluation, aerodynamic analysis was performed using SolidWorks Flow Simulation. The baseline CAD geometry was compared with the refined geometry after VR- and CFD-informed modifications (Figure 5). A mesh convergence study was conducted to ensure the numerical reliability and grid independence of the CFD simulation results. Three mesh configurations (coarse, medium, and fine) with progressively increased mesh densities were evaluated using the drag coefficient (Cd) as the primary validation parameter. The final analysis configuration employed a base mesh dimension in the X, Y, and Z directions and consisted of 3,435 fluid cells, 1,782 solid cells, and 1,578 partial cells.

Grid independence was verified by gradually refining the mesh density and comparing the resulting Cd values among the tested configurations. The minimal variation in Cd values between the medium and fine meshes indicated that further mesh refinement did not significantly affect the simulation outputs. Therefore, the selected mesh configuration was considered sufficiently independent and numerically stable for the final aerodynamic analysis. At a test speed of 80 km/h, the optimal mesh produced a Cd value of 0.404, which falls within the expected drag coefficient range (0.35-0.50) for compact utility-oriented EV bodies.

The reliability of the model was further demonstrated through the precise calculation of aerodynamic forces, including a drag force of 317.519 N and a lift force of 231.367 N, indicating a balanced aerodynamic performance. Compared to the baseline configuration, simulations using the validated independent mesh recorded a 14.6% reduction in peak frontal pressure and an 11.2% improvement in wake stability, confirming the aerodynamic benefits of the refined vehicle geometry. The results indicate that changes to windshield inclination and front-cabin curvature improved both forward visibility and aerodynamic pressure distribution, providing initial evidence for ergonomic-aerodynamic co-optimization (Table 1).

Table 1. Aerodynamic performance results

Figure 6. 3D-printed 1:10 vehicle model

Following digital validation, rapid prototyping was employed to produce both scale models and full-scale mockups of body parts. A 1:10 exterior model (Figure 6) was fabricated to confirm proportions, surface coherence, and aesthetic compatibility. More importantly, a full-scale dashboard mockup (Figure 7) and door panel (Figure 8) enabled physical assessment of driver interaction, control reachability, and spatial ergonomics. These physical evaluations validated many of the earlier VR-driven refinements, demonstrating strong alignment between digital and physical environments. The mockup stage also revealed manufacturing considerations such as panel joinery requirements, material thickness feasibility, and assembly tolerances, leading to additional refinements in CAD geometry and final shop drawings.

Figure 7. Full-scale 3D-printed dashboard mockup (1:1)

The integration of these results demonstrates the strength of the iterative workflow. AI outputs informed CAD geometry; CAD informed VR ergonomics and CFD analyses; VR and CFD results guided further geometric refinement; and rapid prototyping provided tangible validation of usability and manufacturability. This continuous feedback loop reduced avoidable design errors by identifying ergonomic and airflow problems before final fabrication. The process also contributed to the registration of an IPR-protected underfloor cargo innovation for electric pickup-style vehicles, recorded as IPR No. S00202511613. Collaboration with PT Spora Teknika Indonesia ensured that the design remained aligned with local fabrication capacity, assembly constraints, and progression toward higher TRL levels.

The integrated workflow proved highly effective in achieving a refined, ergonomic, aerodynamically optimized, and manufacturable cabin design for a multi-utility EV. The discussion underscores that the convergence of AI, CAD, VR, CFD, and rapid prototyping not only enhances design precision but also offers a scalable model for accelerating EV development in emerging markets, where cost efficiency, iterative speed, and industrial capability are critical constraints.

Figure 8. Full-scale 3D-printed door panel mockup (1:1)

This study demonstrates that an integrated digital workflow, combining AI-assisted generative design, parametric CAD modelling, VR-based ergonomic evaluation, engineering simulation, aerodynamic CFD analysis, and rapid prototyping, provides a highly effective and scalable approach for accelerating the design and development of a multi-utility EV cabin. By linking conceptual, ergonomic, structural, and manufacturing considerations into a continuous iterative cycle, the workflow enables designers to rapidly refine cabin geometry, improve ergonomic comfort, and enhance aerodynamic performance while minimizing design errors and reducing development time. The results illustrate that AI-generated concepts significantly broaden the design space, CAD models ensure dimensional and engineering precision, VR simulations enable early-stage human-machine interaction assessment, and CFD analysis identifies critical aerodynamic issues requiring geometric refinement. The final stage of rapid prototyping validates the digital results, confirming that the integrated workflow produces cabin designs that are not only functional and comfortable but also technically feasible for industrial fabrication.

The findings further indicate that this methodology effectively bridges the common disconnect between digital design environments and physical manufacturing realities, particularly in emerging economies where technological resources and prototyping capabilities are often limited. The successful development of full-scale dashboard mockups, scale models, and an IPR -protected underfloor cargo system demonstrates the workflow’s potential to support innovation pathways from early ideation to near-commercial readiness. The collaboration with PT Spora Teknika Indonesia confirms that the workflow aligns well with industrial constraints, assembly methods, and fabrication practices, positioning the project for advancement along the TRL spectrum as outlined in the research roadmap.

In broader terms, this research contributes to EV design literature by presenting a unified methodology that integrates aesthetic, ergonomic, structural, and aerodynamic analyses within a single, seamless development framework. It offers new insights into the use of AI, VR, and rapid prototyping as complementary tools that collectively enhance design efficiency and user-centered outcomes. For practitioners, the workflow provides a practical model that can help domestic EV manufacturers shorten development cycles, reduce prototyping costs, and improve design accuracy, factors that are critical for increasing competitiveness in the rapidly evolving global EV market.

The study shows that digital integration is not merely a technological trend but a strategic necessity for fostering innovation, improving engineering quality, and strengthening industrial capability in EV development. Despite the promising findings, this study has several limitations. The VR-based ergonomic evaluation involved only five representative participants, which should be considered a pilot-scale sample rather than a statistically representative population. Although the participants were selected to reflect typical Southeast Asian anthropometric characteristics and urban-driving familiarity, the limited sample size may reduce the statistical power and generalizability of the ergonomic findings across broader user populations. Therefore, future studies should incorporate larger and more diverse participant groups to strengthen the robustness, reliability, and external validity of the ergonomic evaluation results.

The revised evidence clarifies the workflow's validation scope, including AI concept generation, VR-based user evaluation, CFD analysis at 80 km/h, 1:10 and 1:1 rapid prototypes, and IPR-oriented innovation output. However, future work should replace the remaining pilot-scale limitations with larger participant samples, broader anthropometric coverage, wind-tunnel or road-test validation, crashworthiness analysis, energy-management modelling, and formal homologation pathways so that the multi-utility EV platform can advance toward commercialization and broader deployment.